PDA

View Full Version : Basal Haplogroup R classification for 24,000-year-old individual (MA-1), from Malíta



R.Rocca
11-20-2013, 06:49 PM
Here are the details behind the grouping of basal Y-haplogroup R for the Ma'lta sample:


Y chromosome haplogroup of MA-1

Due to low depth-of-coverage of the MA-1 individual (1.5X on 5.8 million bases),
genotyping at each site on the Y chromosome was performed by selecting the allele
with the highest frequency of bases with a base quality of 13 or higher. Additionally,
a multi-fasta file was generated from the variable positions on the Y chromosomes
available from 24 Complete Genomics public genomes. SNPs were filtered for
quality (using VQHIGH as the threshold, as defined by Complete Genomics), with
tri-allelic positions excluded and only those Y chromosome regions determined as
being phylogenetically informative being used2. This yielded a final dataset of 22492
positions. MA-1 Y chromosome data was then included, and MEGA phylogenetic
software was used to construct a Neighbor Joining (NJ) tree with default parameters
(Figure SI 5a). MA-1 is placed as a basal lineage to hg R. Phylogenetically
informative positions and their state in MA-1 were then determined to confirm the
placement of MA-1 on the tree. In the course of this analysis, the original dataset was
severely pruned. Non-informative positions, including those with more than four Ns
in the public dataset, were excluded (633 positions). Moreover, the following
positions were also excluded which were 1) in reference state in all individuals
including MA-1 (7172 positions); 2) N in MA-1 and either N or reference state
among the rest of the individuals (9682 positions); 3) ‘N-ref’ – those with only N or
reference state among all individuals (586 positions) and ‘N-alt’ - positions with
alternative alleles, but difficult to classify (11 positions); 4) reference specific (79
positions); and, 5) recurrent (28 positions). This resulted in 4301 positions being
retained that were classified according to their hg affiliations. Among those
phylogenetically informative positions, 1889 non-N positions were retrieved from
MA-1. When counting from the split of hg DE on the unrooted phylogenetic tree,
MA-1 is determined to be carrying the derived allele in 183 sites and the ancestral
allele in 1706 sites. The position of MA-1 on the phylogenetic tree is established by
the state of the 313 basal mutations separating hgs DE and R, where MA-1 has 143
informative positions. Of these, 138 are in the derived and 5 in the ancestral state,
placing MA-1 as a lineage basal to hg R. With only a few exceptions characterized
below, all other informative positions in MA-1 are in the ancestral state, further
supporting the phylogenetic positioning of MA-1 on the tree.
Among the derived markers in the final dataset only a few (11) mutations were
detected that are likely to be false positives based on the phylogenetic analysis, where
it is assumed that recurrent mutation is less likely than a sequencing error. One
position among the 35 private to MA-1 is characteristic of a distant hg – namely
C3c14. Based on current data, 10 additional phylogenetically non-concordant
positions in MA-1 were found – 1 position for hgs E, G, Q, R1b, R1 each, 2 defining
positions for hg I and 3 private mutations for R1b individuals (shown in red on Figure
SI 5a). Additionally, among the mutations originally excluded (the reference-private
mutations), two positions were found where MA-1 is in derived state.

http://r1b.org/imgs/MA-1_Tree.png


Figure SI 5a An unrooted Neighbor Joining tree of 24 Y chromosomes from the
publicly available Complete Genomics dataset1 and MA-1. The tree is based on
22492 Y chromosome SNPs and constructed with MEGA using default parameters.
MA-1 is ancestral to haplogroup R, based on 5 sites that are in ancestral state
compared to all other individuals belonging to this haplogroup. The phylogenetic
analysis of sites underlying the tree resulted in a pruned dataset of 4301 informative
positions. The derived (d) or ancestral (a) states of called (non-N) sites in the MA-1 Y
chromosome are noted on the edges of the tree. The overall number of informative
sites at those edges is shown in parentheses. For example, “(232) MA-1: 109a, 1d,
122N ” on the hg Q edge means that out of the 232 positions in derived state at this
edge, 109 were ancestral, one derived and 122 N in MA-1.

Jean M
11-20-2013, 06:52 PM
Where is the data from?

[Added] It's OK. I see you announced the paper on another thread. http://www.nature.com/nature/journal/vaop/ncurrent/full/nature12736.html

Joe B
11-20-2013, 06:58 PM
Here are the details behind the grouping of basal Y-haplogroup R for the Ma'lta sample:
MA-1 is placed as a basal lineage to hg R2,4. Phylogenetically
informative positions and their state in MA-1 were then determined to confirm the
placement of MA-1 on the tree.
Based on current data, 10 additional phylogenetically non-concordant
positions in MA-1 were found – 1 position for hgs E, G, Q, R1b, R1 each, 2 defining
positions for hg I and 3 private mutations for R1b individuals (shown in red on Figure
SI 5a).
Please explain! Is it R2 or R1b?

ADW_1981
11-20-2013, 07:12 PM
Please explain! Is it R2 or R1b?

If I understand correctly it looks like a basal branch of R1b1 or R1b, and if I am looking at the tree correctly. Hopefully others can chime in with additional input or correct me why this is not so.

R.Rocca
11-20-2013, 07:12 PM
Please explain! Is it R2 or R1b?

Sorry, when I copy and pasted the test, the references to other sudies went from super-script to normal. It is now fixed...it is basal R, not R2.

parasar
11-20-2013, 07:22 PM
Sorry, when I copy and pasted the test, the references to other sudies went from super-script to normal. It is now fixed...it is basal R, not R2.

Looks R(xR2) right? Pre-R1?

Jean M
11-20-2013, 07:24 PM
Basal R = R*.

R.Rocca
11-20-2013, 07:27 PM
Looks R(xR2) right? Pre-R1?

Correct. It looks like the sample has some false positives in some positions, but there are more than enough positives (derived) to make him R+ and enough negatives (ancestral) to make him R1- and R2-.

parasar
11-20-2013, 07:27 PM
http://www.nature.com/nature/journal/vaop/ncurrent/extref/nature12736-s1.pdf


10 additional phylogenetically non-concordant positions in MA-1 were found – 1 position for hgs E, G, Q, R1b, R1 each, 2 defining
positions for hg I and 3 private mutations for R1b individuals

Do we know why R1 is phylogenetically non-concordant?

R.Rocca
11-20-2013, 07:41 PM
http://www.nature.com/nature/journal/vaop/ncurrent/extref/nature12736-s1.pdf

Do we know why R1 is phylogenetically non-concordant?

If I am reading the image legend correctly, it looks like it is non-concordant because, while a single R1b SNP was derived, all other R1 defining mutations with reads between it and basal R were ancestral.

ADW_1981
11-20-2013, 08:16 PM
Come to think of it, I guess this really shouldn't be a surprise considering the high extinction rates of lineages that this line, nor the basal U line crossed Beringia. All NA/SA lineages are derived from Q1-L53+ if I recall correctly. This is just a single man who likely lived in Siberia. It would be interesting to see if he was around, or where he lived with respect to the R* guy.

R.Rocca
11-20-2013, 09:01 PM
By the way, it is interesting to note that this sample is placed parallel to R1 and R2. Another words, he has enough private mutations to put him in a new parallel branch (future R3?) and likely has no known living male descendants. Would be nice to try to sequence the area to see if that hold true.

parasar
11-20-2013, 09:12 PM
By the way, it is interesting to note that this sample is placed parallel to R1 and R2. Another words, he has enough private mutations to put him in a new parallel branch (future R3?) and likely has no known living male descendants. Would be nice to try to sequence the area to see if that hold true.

Is it possible he is not parallel to R2, but parallel to R-1 only, i.e. pre-R1(xR2) with the branching occurring after the one mutation he shares with R1.

RCO
11-20-2013, 09:12 PM
Not R3, but the Mal'ta should become R2, R1=R1a, R2 = R1b, R1a = R1a1 and R1b=R1a2
We are used to new basal branches restructuring the J1 tree !

Adversary
11-20-2013, 09:54 PM
Is it possible he is not parallel to R2, but parallel to R-1 only, i.e. pre-R1(xR2) with the branching occurring after the one mutation he shares with R1.
How is that possible? He lacks 5 mutations that are shared between R1 and R2. How could it be meaningful that he happens to share just one mutation with downstream R1, and another with R1b (also E, G, Q, and he even shares 2 defining mutations with I)?

Rathna
11-20-2013, 10:02 PM
How is that possible? He lacks 5 mutations that are shared between R1 and R2. How could it be meaningful that he happens to share just one mutation with downstream R1, and another with R1b (also E, G, Q, and he even shares 2 defining mutations with I)?

As in the Y there isn’t the Heteroplasmy like in the mtDNA, we have to think that the mutation as to R1b is an independent one, different from that occurred in our line. The mutations that lack (5) are overwhelming, I think. Then I opt for an R*, and of course not of the survived (our ) line.

parasar
11-20-2013, 10:08 PM
How is that possible? He lacks 5 mutations that are shared between R1 and R2. How could it be meaningful that he happens to share just one mutation with downstream R1, and another with R1b (also E, G, Q, and he even shares 2 defining mutations with I)?

That makes sense.
Do we know which are ones shared between R1 and R2 that the Mal'ta boy does not have? (M207/Page37/UTY2, P224, P227, P229/PF6019, P232, P280, P285, S4, S9)
Thanks.

alan
11-20-2013, 10:13 PM
Do his private SNPs not by their existence give and indication of how long he lived after R came into existence? He dates to 22000BC but R is usually thought to be a few thousand years older than that.

parasar
11-20-2013, 10:17 PM
Do his private SNPs not by their existence give and indication of how long he lived after R came into existence? He dates to 22000BC but R is usually thought to be a few thousand years older than that.

I think the privates + non-concordant could give us a ballpark minimum separation.

alan
11-20-2013, 10:21 PM
Anyone any good at using SNP counts to work out timespans? How long after R does this guy seem to have lived?

R.Rocca
11-20-2013, 11:46 PM
Do his private SNPs not by their existence give and indication of how long he lived after R came into existence? He dates to 22000BC but R is usually thought to be a few thousand years older than that.

I don't think it is possible since they only did a 1x scan. By comparison, 1000 Genomes was a 2x-4x scan, Full Genomes (50x) and Big-Y (60x), so there are likely a lot of missing data.

alan
11-21-2013, 12:17 AM
Oh well. Its still interesting that a number of private SNPs were picked up even in a low resolution Scan. I dont suppose there are any consistent ratios of SNPs picked up in scan like this compared to in a very fine grained scan?

Karafet estimated R at 26,800 (19,900 - 34,300) years ago. This datapoint of 24000 years ago for someone some time after R occurred at least seems to show that R falls somewhere between the centre or older half of the age estimate range.




I don't think it is possible since they only did a 1x scan. By comparison, 1000 Genomes was a 2x-4x scan, Full Genomes (50x) and Big-Y (60x), so there are likely a lot of missing data.

alan
11-21-2013, 12:38 AM
Has anyone actually read the full paper?

R.Rocca
11-21-2013, 12:40 AM
Has anyone actually read the full paper?

Yes, it is a quick read, but quite fascinating when one thinks of what is now possible. The meat of it is in the supplementary information though.

parasar
11-21-2013, 01:16 AM
As in the Y there isn’t the Heteroplasmy like in the mtDNA, we have to think that the mutation as to R1b is an independent one, different from that occurred in our line. The mutations that lack (5) are overwhelming, I think. Then I opt for an R*, and of course not of the survived (our ) line.

The way I had initially read it, as between DE and R, I had had thought it was not that significant, but at R* itself (as depicted in the figure) it becomes quite significant.


The position of MA-1 on the phylogenetic tree is established by
the state of the 313 basal mutations separating hgs DE and R, where MA-1 has 143
informative positions. Of these, 138 are in the derived and 5 in the ancestral state,
placing MA-1 as a lineage basal to hg R.

ADW_1981
11-21-2013, 01:33 AM
I am comparing this study to the following which analyzes the diversity of Q1a3a-M3 among Native Americans. From the YDNA perspective it looks like the Q1a3 population was sourced from an area between Mongolia and the Kamchatka peninsula if we examine the YDNA. However, this is causing me to think deeper and generating more questions than answers.

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0071390


EDIT: I just noted that the R*, U* remains are in central Siberia as opposed to NE Siberia where it seems the Amerindian founders came from. This presents one of two possibilities in my mind.

a) a Caucasoid"ish" population of Q1a3 was already present in NE Siberia and then absorbed a northward moving "East Asian" population consisting of a combination of mtDNA A, B, C, D, and YDNA C3 lineages.
b) a Caucasoid"ish" population of Q1a3 and possibly R* moved from Baikal to NE Siberia and absorbed an already present "East Asian" population consisting of a combination of mtDNA A, B, C, D, and YDNA C3 lineages - losing mtDNA lineages such as U* along the way.

If even one of my scenarios were correct, I wonder what the implications are for seemingly "ancient" Q1a3(xL53, L54) populations in Europe which are mostly restricted to places like the UK, Norway, and Sweden. Would they have arrived before the R1(b/a) ancestor or simultaneously as a minority lineage.

alan
11-21-2013, 01:36 AM
What about the Afontova Gora person? The comments on the web are a bit vague. They seemed to be implying autosomal continuity with the Malta guy. If that is the case then that is telling us something very interesting indeed as the area was abandoned in the height of the LGM by the middle upper Palaeolithic culture we see at Malta for quite a few millennia before the microblade groups emerged somewhere like Altai or east Asia and occupied the empty area and brought the sort of culture we see at Afontova Gora. That would imply that the microblade technology emerged among some of the middle Palaeolithic Malta type Siberians who had headed into refugia to the south and developed the technology there before expanding north again.

Another aspect is that the post-LGM late upper palaeolithic guy at Afontova Gora had previously been identified tentatively as Mongoloid. So, perhaps the mixing between Malta type groups and east Asians happened during the LGM when people were squeezed together into regugia. Obviously that would fit quite well the sort of mix that headed to America.

parasar
11-21-2013, 04:18 AM
How is that possible? He lacks 5 mutations that are shared between R1 and R2. How could it be meaningful that he happens to share just one mutation with downstream R1, and another with R1b (also E, G, Q, and he even shares 2 defining mutations with I)?

Doesn't the fact that R1 is immediately downstream make a difference?

parasar
11-21-2013, 04:31 AM
...

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0071390


EDIT: I just noted that the R*, U* remains are in central Siberia as opposed to NE Siberia where it seems the Amerindian founders came from ...

From the link you posted:

These observations are in agreement with the hypothesis that the Native American ancestral population comes from somewhere in southern Siberia, between the Altai Mountains (which acted as a genetic and cultural wall) and the Lake Baikal region

The Mal'ta R, U is from the eastern end of the region.

parasar
11-21-2013, 04:36 AM
What about the Afontova Gora person? The comments on the web are a bit vague. They seemed to be implying autosomal continuity with the Malta guy. If that is the case then that is telling us something very interesting indeed as the area was abandoned in the height of the LGM by the middle upper Palaeolithic culture we see at Malta for quite a few millennia before the microblade groups emerged somewhere like Altai or east Asia and occupied the empty area and brought the sort of culture we see at Afontova Gora. That would imply that the microblade technology emerged among some of the middle Palaeolithic Malta type Siberians who had headed into refugia to the south and developed the technology there before expanding north again.

Another aspect is that the post-LGM late upper palaeolithic guy at Afontova Gora had previously been identified tentatively as Mongoloid. So, perhaps the mixing between Malta type groups and east Asians happened during the LGM when people were squeezed together into regugia. Obviously that would fit quite well the sort of mix that headed to America.

If that were the case, Raghavan et. al. would have noted the change.

ADW_1981
11-21-2013, 02:31 PM
From the link you posted:


The Mal'ta R, U is from the eastern end of the region.

Yes, I would argue on the male side this seems to be the case, I guess the R* lineages just died out and/or never moved east. the Q1a3a-M3 haplotypes which represents 99% of that paper's sample seems to coincide with data between the range of Siberia and the Kamchatka peninsula.

Scrolling back in the phylogeny we have Q1a-MEH2 which also has a wide dispersal - found in Kamchatka, the Paleo-Eskimo (5,500 ybp from Greenland), and ...Afghan Pashtuns. R* is allegedly found among the Pashtuns as well. I would have normally thought this was a bad result, but now I am re-considering after looking at the relationship between Q-R. Perhaps we can finally nail down the origin of Q-R to one of the "Stan" countries. For whatever reason this group had spread into Siberia and further east. It might have happened that "East Asians" began spreading north from NE China/Korea/Japan area < 17,000 ago and may have rekindled a dying northern population.

parasar
11-21-2013, 03:01 PM
Yes, I would argue on the male side this seems to be the case, I guess the R* lineages just died out and/or never moved east. the Q1a3a-M3 haplotypes which represents 99% of that paper's sample seems to coincide with data between the range of Siberia and the Kamchatka peninsula.

Scrolling back in the phylogeny we have Q1a-MEH2 which also has a wide dispersal - found in Kamchatka, the Paleo-Eskimo (5,500 ybp from Greenland), and ...Afghan Pashtuns. R* is allegedly found among the Pashtuns as well. I would have normally thought this was a bad result, but now I am re-considering after looking at the relationship between Q-R. Perhaps we can finally nail down the origin of Q-R to one of the "Stan" countries. For whatever reason this group had spread into Siberia and further east. It might have happened that "East Asians" began spreading north from NE China/Korea/Japan area < 17,000 ago and may have rekindled a dying northern population.

The high number of P(xQ-M242, xR-207) reported from southern India tells me that differentiation to R may have occurred there.

R.Rocca
11-21-2013, 03:16 PM
The high number of P(xQ-M242, xR-207) reported from southern India tells me that differentiation to R may have occurred there.

That depends. If all of those P(xQ-M242, xR-207) form part of a single yet-to-be discovered subclade of P, then its existence in southern India is no more significant than R in Europe or Q in the Americas.

parasar
11-21-2013, 03:33 PM
That depends. If all of those P(xQ-M242, xR-207) form part of a single yet-to-be discovered subclade of P, then its existence in southern India is no more significant than R in Europe or Q in the Americas.

They have M207 (M479 status not known) and M242 too. If as Hammer posits that the trajectory of P-R is from SE Asia to Europe, southern India may have been the mid-point.

They may well be under one unifying clade, though from the STR data it is difficult to predict.
DYS389I DYS389II DYS390 DYS456 DYS19 DYS458 DYS437 DYS438 DYS448 DYSGATAH4 DYS391 DYS392 DYS393 DYS439 DYS635 DYS388 DYS426

P-M45(xM207, xM242)
PLY HTF3 HTF P* P-M45 13 16 22 16 13 17 15 11
THD HTC2 HTC P* P-M45 13 17 25 17 16 15 14 11
NDC DLF5 DLF P* P-M45 14 16 22 15 13 19 14 11
VLR AW1 AW P* P-M45 14 18 24 15 16 16 14 11
VLR AW1 AW P* P-M45 14 18 24 15 16 16 14 11
VLR AW1 AW P* P-M45 14 16 22 15 13 17 14 11
BHC BRH2 BRH P* P-M45 13 18 25 16 15 14 14 11

Eyeballing, M207 diversity appears higher, as if there are a number of R lines:

R-M207 (M479 status not known)
Maravar
R* R-M207 14 20 23 16 14 18 16 11 19 12 10 10 15 12 26 12 11
R* R-M207 14 19 24 15 14 15 14 10 19 12 10 11 14 12 18 12 11
R* R-M207 14 17 24 15 15 17 14 11 20 12 11 11 13 10 23 12 11
R* R-M207 14 18 22 15 15 17 14 10 19 11 11 11 13 11 20 13 11
R* R-M207 12 19 25 16 12 16 14 11 20 12 10 11 13 10 23 12 12
R* R-M207 14 16 23 16 14 16 15 12 19 13 11 13 12 12 23 12 12
Paniya
R* R-M207 15 17 23 15 15 18 15 11 19 12 10 10 14 11 25 13 12
Irula
R* R-M207 14 16 24 15 15 17 16 11 19 12 10 10 14 12 25 12 12
Kanikaran
R* R-M207 13 17 25 15 17 15 14 11 20 12 11 11 13 10 23 12 12
Kota
R* R-M207 13 18 25 16 15 14 14 11 20 12 11 11 13 10 23 12 10
R* R-M207 14 17 25 15 15 16 14 11 20 12 10 11 13 10 23 12
R* R-M207 13 16 23 15 14 17 15 9 21 12 10 13 12 24 15 11
Kattunaickan
R* R-M207 12 16 23 13 15 18 14 10 19 11 10 12 13 11 25 14 11
R* R-M207 13 15 22 17 14 20 16 11 19 12 10 14 12 24
Parayar
R* R-M207 14 17 24 14 15 16 14 11 20 13 10 11 13 11 24 12 12
R* R-M207 13 17 25 16 15 15 14 10 20 12 10 11 13 10 23 12 12
R* R-M207 13 16 23 16 14 18 16 11 19 12 10 10 14 10 24 12 12
Pallar
R* R-M207 13 17 25 16 16 15 14 11 20 12 10 11 13 10 23 12 10
Paravar
R* R-M207 14 14 23 15 15 17 15 11 19 12 10 10 13 12 24 12 12
Yadhava
R* R-M207 13 16 24 14 15 14 14 11 18 12 10 13 13 11 26 13 12
Vanniyar
R* R-M207 13 17 24 15 13 16 15 12 19 13 10 13 12 12 23 12 11
R* R-M207 14 16 23 15 14 20 16 11 19 10 10 10 14 10 24 12
Cape Nadar
R* R-M207 14 16 23 15 14 16 14 11 19 11 10 15 12 12 26 12 12
Valayar
R* R-M207 14 16 22 15 15 16 15 10 18 12 10 11 13 12 21
Mukkuvar
R* R-M207 13 17 25 15 14 18 15 12 19 11 10 13 12 13 23 12 12
R* R-M207 13 23 17 17 15 12 19 12 10 12 11 23 12 12
Brahacharanam
R* R-M207 13 17 22 17 18 16 15 10 19 12 10 11 13 13 21 12 12
R* R-M207 14 16 23 16 15 16 15 12 19 13 11 13 12 13 23 12 12
R* R-M207 13 18 24 16 16 15 14 11 20 13 10 13 10 23 12 10
R* R-M207 14 16 23 17 15 16 15 12 19 13 11 12 13 23 12 11
VadamaBrahmin
R* R-M207 13 15 24 15 15 19 16 11 19 12 10 10 13 10 24 12 11
R* R-M207 14 16 24 15 14 18 16 12 19 11 10 10 13 11 25 12 12
R* R-M207 13 18 24 15 16 14 16 11 19 11 10 10 14 10 26 12 12
R* R-M207 13 18 23 15 14 17 15 11 19 12 10 10 14 10 13 12

Generalissimo
11-21-2013, 11:53 PM
I'm betting that the genomes from the Sungir site in Russia, from near Moscow, will also show a basal R lineage, mtDNA U, and come out like a mix of North Euros, South or West Asians and Amerindians in terms of genome-wide structure.

This is all very cool, and maybe surprising at first, but it makes a lot of sense; the R and Q link between Europeans and Amerindians; the R and autosomal North Euro + South Asian links between Europeans and South Asians; the close genetic (Fst) relationship between the ADMIXTURE clusters modal in Northern/Eastern Europeans and West Asians, but a massive gap on PCA plots between Eastern Europe and the Caucasus.

So the Upper Paleolithic R1 and U4, U5 and U2 hunter-gatherers of the Eastern European mammoth steppe were North Euro/South or West Asian/Amerindian-like, and then came the Sardinian-like Oetzis from the west, or at least some of their DNA filtered through during and after the Neolithic, and we got modern Eastern Europeans.

But I'd say the Siberian-like Uralics joined the party later, and their component (+ N1c1) wasn't a feature of the aforementioned North Euro/South or West Asian/Amerindian-like European foragers. Pale yellow component below...

http://img43.imageshack.us/img43/2810/hhx4.png

parasar
11-22-2013, 01:22 AM
I'm betting that the genomes from the Sungir site in Russia, from near Moscow, will also show a basal R lineage, mtDNA U, and come out like a mix of North Euros, South or West Asians and Amerindians in terms of genome-wide structure.
...

I wouldn't bet against an R either.
Kostenki 14 too I think was P or R.

alan
11-22-2013, 01:27 AM
That is a big question. I dont know if it clear that these Mal'ta types were the same as the Gravettian groups of eastern Europe. They had something in common but a lot different too. For a start, the Mal'ta boy was culturally one of the very last of what was the middle upper palaeolithic culture of Siberia. As far as I recall, this culture commenced around 30000BC and is a little older than Gravettian. There were also late Aurignacian groups in European Russia too though.

So I find it hard to work out if the Mal'ta boy was some sort of peripheral outpost and perhaps hybrid of an eastern European group or whether the Mal'ta group simply were Siberian and partly moved west during the LGM, perhaps making it into the eastern fringes of Europe. I just cannot work it out from either the genetics or the archaeology. Based on very old branchings it is easier to see R heading south of the central Asian desert in the LGM. On the other hand following the mammoth steppe directly west at a higher latitude would have been more in line with the sort of hunting practices of the culture the Mal'ta boy lived in. To further confuse the issue, if the Afantova Gora guy has similar genetics then it is possible that they headed straight south the relatively short distance into Altai or the like. It very hard to guess.


I'm betting that the genomes from the Sungir site in Russia, from near Moscow, will also show a basal R lineage, mtDNA U, and come out like a mix of North Euros, South or West Asians and Amerindians in terms of genome-wide structure.

This is all very cool, and maybe surprising at first, but it makes a lot of sense; the R and Q link between Europeans and Amerindians; the R and autosomal North Euro + South Asian links between Europeans and South Asians; the close genetic (Fst) relationship between the ADMIXTURE clusters modal in Northern/Eastern Europeans and West Asians, but a massive gap on PCA plots between Eastern Europe and the Caucasus.

So the Upper Paleolithic R1 and U4, U5 and U2 hunter-gatherers of the Eastern European mammoth steppe were North Euro/South or West Asian/Amerindian-like, and then came the Sardinian-like Oetzis from the west, or at least some of their DNA filtered through during and after the Neolithic, and we got modern Eastern Europeans.

But I'd say the Siberian-like Uralics joined the party later, and their component (+ N1c1) wasn't a feature of the aforementioned North Euro/South or West Asian/Amerindian-like European foragers. Pale yellow component below...

http://img43.imageshack.us/img43/2810/hhx4.png

alan
11-22-2013, 01:33 AM
He came from a pre-Gravettian level. I tend to think that IJ is related to Gravettian. I think the culture that the Malta boy was part of was pre-gravettian. So you may have a point that he could have been from some sort of P derived line.


I wouldn't bet against an R either.
Kostenki 14 too I think was P or R.

alan
11-22-2013, 08:31 AM
Sorry I meant central Asian desert - changed now.

Another thing is of course R2. It may have come into existence just after the Malta boy so its another indicator of what fairly close cousins of the Malta boy did in the LGM. They clearly seem to have taken the route south. I suspect the R* line that led to R2 left Siberia before the Malta boy was born. It just seems to me that in order to head south-west like that they had to beat the formation of the central Asian LGM deserts which drove a barrier between the mammoth steppe tundra belt in the north and the area south of the Desert along the Hindu Kush to Iran latitude to the south. I think R2 is compelling evidence that some did that. I feel that any R* that remained in Siberia as late as the Malta boy probably no longer had the option of taking that route.

Anyone who stayed that late would either have to head west along the mammoth steppe or down into Altai/Dzungaria which may also have allowed passage to the Pacific. Given the lack of R*, R2 etc in that area it seems that this route was not taken in any numbers by R* but it may have been taken by Q. I personally suspect the Afontova Gora burial may be Q. It is nearly 10000 years younger than the Mal'ta guy and has a completely different technology which probably arose in a refuge around Altai during the LGM where that technology is old. The real expansion of this microblade group was post-LGM and the capacity to settle in colder areas that previously settled was part and parcel of this new technology, as was extreme mobility compared to the relatively settled Mal'ta type middle upper palaeolithic macroblade cultures with their elaborate base camps.

The fact Afontova has some similarity in genetics to Mal'ta does indicate to me that some of these pre-LGM siberians did make the shorter move to Altai before re-expanding in post-LGM times with a different technology. This doesnt seem to have included R but I would strongly suspect Q was involved and this has something to do with the settlement of the Americas.



That is a big question. I dont know if it clear that these Mal'ta types were the same as the Gravettian groups of eastern Europe. They had something in common but a lot different too. For a start, the Mal'ta boy was culturally one of the very last of what was the middle upper palaeolithic culture of Siberia. As far as I recall, this culture commenced around 30000BC and is a little older than Gravettian. There were also late Aurignacian groups in European Russia too though.

So I find it hard to work out if the Mal'ta boy was some sort of peripheral outpost and perhaps hybrid of an eastern European group or whether the Mal'ta group simply were Siberian and partly moved west during the LGM, perhaps making it into the eastern fringes of Europe. I just cannot work it out from either the genetics or the archaeology. Based on very old branchings it is easier to see R heading south of the central Asian desert in the LGM. On the other hand following the mammoth steppe directly west at a higher latitude would have been more in line with the sort of hunting practices of the culture the Mal'ta boy lived in. To further confuse the issue, if the Afantova Gora guy has similar genetics then it is possible that they headed straight south the relatively short distance into Altai or the like. It very hard to guess.

alan
11-22-2013, 09:32 AM
In general, while I think the IJ group links to Gravettians, I suspect the roots of P are in pre-Gravettian groups who headed east in the KxLT/MNOP group well before Gravettian. P was perhaps the one branch of that movement that took the northern route. There are early upper palaeolithic groups even in Siberia from around 40000BC which perhaps could correspond to early P. They also seem to have been driven south for a time by climate fluctuations before the Siberian middle upper palaeolithic cultures formed c. 30000BC, a date with might broadly date to the beginings of R or perhaps an immediately ancestral P* group. Whether they were essentially the same people re-expanding north with technological adjustments or another fresh wave is unclear to me.

If P is the odd man out branch that was the only one to take a northern route and ended up in Siberia and adjacent then it is interesting to speculate where exactly it split. We must remember than modern patterns may reflect the impact of the LGM and that central Asia was a desert and deserted meaning that examples there could not have been in that location during the LGM. i have posted before that I have read that the only likely pass from south Asia between Afghanistan and the Pacific is Dzungaria. So, that would suggest limited options for P to have split and taken a different northern route to the rest of MNOP. I have major doubts that we can look to the Pacific area for P entering the northern area. So, that really just leaves the Dzungarian gate or a split much further west around Afghanistan or Iran.

II have got to say I find a route north into Siberia from Dzungaria very unlikely as this was essentially a route from Mongolia or north Pacificeast Asia into the east end of the central Asia steppe. It also hard to square with modern survival of the P paragroup and descendant lineage. So, I would feel that this points to P diverging from the MNOP group around Iran and heading east through north central Asia. The date for this would appear to place it in the early upper palaeolithic c. 40000BC. That sort of model would nicely explain the deep separation from potentially Gravettian associated I which would represent a later expansion from the Levant through Anatolia and the Caucasus into Europe.

I have read the recent suggestion of some sort of really complex retreat of K descended lineages back west. I presume that is based on the paragroup K being Oceanian and more eastern compared to many of its descendant groups. I dont know what to make of that. Perhaps that is in some way related to the apparently climatically driven hiatus between early upper palaeolithic groups and middle upper palaeolithic groups in some areas including Siberia.

alan
11-22-2013, 01:50 PM
I found Eurologist's post on Dienekes blog interesting

Again, some of the wording in the paper is very unfortunate. We should not forget that both Amerindians and Europeans derive from sister populations of Mal'ta boy - and not his exact group (both his y-DNA and mtDNA are extinct yet derived, and to both sides different haplogroups were present (different U subgroups and IJ and R1 to the west, and Q and X to the East). To both sides the impact of these people - all apparently originating from the NW subcontinent - was actually earlier than Mal'ta boy: the Gravettian started >~32 kya, and Siberia was settled west to east 37ky to 35kya.

So, those ~10ky of drift means that the full contribution of these (outer) populations to both Europeans and Amerindians is likely underestimated in this paper (because it actually derives from populations ~10ky removed from Mal'ta boy).


Mal'ta also had it seems from K=9 more south Asian than European ancestry.

Barak,

Of course - but that just appears like that because you are defining an ancient population via modern components. All Paleolithic and Mesolithic Europeans, northern* West Asians, and non-East-Asian-admixed Siberians very likely originate from the NW part of the subcontinent. Extant Europeans have little S Asian now, because both had 35-45ky to drift apart to make them into exactly these identifiable components now, in the first place.

* Ancient and modern SW Asian may have/ have admixture from Africa and any possible surviving "Arabian Gulf Oasis" population.

parasar
11-22-2013, 03:09 PM
I found Eurologist's post on Dienekes blog interesting

Again, some of the wording in the paper is very unfortunate. We should not forget that both Amerindians and Europeans derive from sister populations of Mal'ta boy - and not his exact group (both his y-DNA and mtDNA are extinct yet derived, and to both sides different haplogroups were present (different U subgroups and IJ and R1 to the west, and Q and X to the East). To both sides the impact of these people - all apparently originating from the NW subcontinent - was actually earlier than Mal'ta boy: the Gravettian started >~32 kya, and Siberia was settled west to east 37ky to 35kya.

So, those ~10ky of drift means that the full contribution of these (outer) populations to both Europeans and Amerindians is likely underestimated in this paper (because it actually derives from populations ~10ky removed from Mal'ta boy).


Mal'ta also had it seems from K=9 more south Asian than European ancestry.

Barak,

Of course - but that just appears like that because you are defining an ancient population via modern components. All Paleolithic and Mesolithic Europeans, northern* West Asians, and non-East-Asian-admixed Siberians very likely originate from the NW part of the subcontinent. Extant Europeans have little S Asian now, because both had 35-45ky to drift apart to make them into exactly these identifiable components now, in the first place.

* Ancient and modern SW Asian may have/ have admixture from Africa and any possible surviving "Arabian Gulf Oasis" population.

That part is critical. Admixture components are not necessarily only components of but were also components from at some time in the past.

alan
11-22-2013, 03:57 PM
I think his point about it being R and U against the nearest new world thing being Q and X is important. The settlers in America were P branch cousins rather than descendants of Ma''ta. I understand Q is younger than R and may have still been P* at the time of the Mal'ta boy. Is there any hint of autosomal similarity to the Mal'ta component among P paragroups. Or is it the result of an isolated group that contained R and Q or its immediate ancestor?



That part is critical. Admixture components are not necessarily only components of but were also components from at some time in the past.

Jean M
11-22-2013, 05:42 PM
I found Eurologist's post on Dienekes blog interesting

Again, some of the wording in the paper is very unfortunate. We should not forget that both Amerindians and Europeans derive from sister populations of Mal'ta boy - and not his exact group

Roberta Estes makes the same point in her post today: Native American Gene Flow – Europe?, Asia and the Americas
http://dna-explained.com/2013/11/22/native-american-gene-flow-europe-asia-and-the-americas/


Pre-release information from the paper, “Upper Palaeolithic Siberian genome reveals dual ancestry of Native Americans” which included results and analysis of DNA sequencing of 24,000 year old skeletal remains of a 4 year old Siberian boy caused quite a stir. Unfortunately, it was also misconstrued and incorrectly extrapolated in some articles. Some people misunderstood, either unintentionally or intentionally, and suggested that people with haplogroups U and R are Native American. That is not what either the prerelease or the paper itself says.

She explains beautifully.


Within the paper, the authors also compare the MA-1 sequence to that of another 40,000 year old individual from Tianyuan Cave, China whose genome has been partially sequenced. This Chinese individual has been shown to be ancestral to both modern-day Asians and Native Americans. This comparison was particularly useful, because it showed that MA-1 is not closely related to the Tianyuan Cave individual, and is more closely related to Native Americans. This means that MA-1’s line and Tianyuan Cave’s line had not yet met and admixed into the population that would become the Native Americans. That occurred sometime later than 24,000 years ago and probably before crossing Beringia into North America sometime between about 18,000 and 20,000 years ago.


The full post is well worth reading - clear and well-illustrated.

parasar
11-22-2013, 07:44 PM
I think his point about it being R and U against the nearest new world thing being Q and X is important. The settlers in America were P branch cousins rather than descendants of Ma''ta. I understand Q is younger than R and may have still been P* at the time of the Mal'ta boy. Is there any hint of autosomal similarity to the Mal'ta component among P paragroups. Or is it the result of an isolated group that contained R and Q or its immediate ancestor?

The paper makes it quite clear:

and near the root of most Native American lineages5. Similarly, we find autosomal evidence that MA-1 is basal to modern-day western Eurasians and genetically closely related to modern-day Native Americans

parasar
11-22-2013, 07:52 PM
Roberta Estes makes the same point in her post today: Native American Gene Flow – Europe?, Asia and the Americas
http://dna-explained.com/2013/11/22/native-american-gene-flow-europe-asia-and-the-americas/



She explains beautifully.



The full post is well worth reading - clear and well-illustrated.

Re: "This means that MA-1’s line and Tianyuan Cave’s line had not yet met and admixed into the population that would become the Native Americans. That occurred sometime later than 24,000 years ago and probably before crossing Beringia into North America sometime between about 18,000 and 20,000 years ago."

The paper's position:


Furthermore, we estimate that 14 to 38% of Native American ancestry may originate through gene flow from this ancient population. This is likely to have occurred after the divergence of Native American ancestors from east Asian ancestors, but before the diversification of Native American populations in the New World.

Jean,

Has anyone reported on the Blood Group of the Mal'ta boy?

Thanks.

parasar
08-07-2014, 02:12 AM
Re: "This means that MA-1’s line and Tianyuan Cave’s line had not yet met and admixed into the population that would become the Native Americans. That occurred sometime later than 24,000 years ago and probably before crossing Beringia into North America sometime between about 18,000 and 20,000 years ago."

The paper's position:



Jean,

Has anyone reported on the Blood Group of the Mal'ta boy?

Thanks.

Anyone?
Do we even have the information to determine?